THE FERTILISATION OF SALYIA AND OTHER FLOWERS. 265 
.front at the upper part, the style is necessarily so bent round as 
to point towards the opening of the corolla tube. When the 
flower first expands, and for some time afterwards, the tip of the 
style is seen just projecting from the highest point of the front 
opening of the hood (fig. 2). The extremity of the style is bifid, 
and the stigmas occupy the internal surface of the terminal 
divisions. When the flower is freshly opened these terminal 
divisions are not separated, but are in close adherence to each 
other by their inner or stigmatic surfaces ; neither are the divi- 
sions, nor the style, as a whole, nearly so long as they after- 
wards become. In fact, the stigmas are not yet ripe, whereas 
the anthers are. This difference in the period of maturity of 
anther and stigma, in itself is a security against self-fertilisation, 
but there are additional safeguards. The style, as already said, 
is prolonged much beyond the upper anther cells, and thus it is 
impossible for these cells, when they move in and out of the 
hood, to strike upon the stigma. This is still further prevented 
by the anther cells being not only below but in front of the 
stigma, so that their motion occurs without passing by the 
style at all, and again by the dehiscence of the anther cells 
being on the side turned away from the style, namely, in front. 
The bee, in its exit from the flower, does not touch the stigma 
while it is in this state, as anyone will easily see if he watches 
a plant for a time ; for the bee retires as he entered, along the 
lower lip, and the width of the opening is too wide for the back 
of the bee to come into contact with the end of the style. Even 
were this contact to occur by accident, it would probably be of no 
effect ; for the stigma is at this period immature, and the viscid 
internal surfaces of the terminal divisions of the style are in 
contact with each other and not exposed. 
The bee therefore flies off with its back smeared with pollen. 
Should it fly to another blossom in the same condition as that 
just described, its cargo of pollen will produce no effect ; all 
that will happen will be that an addition will be made to the 
pollen on its back. But the bee will inevitably soon visit a 
blossom in a more advanced condition than was that from 
which its pollen was derived. Here it will find the stigma in a 
different position. After the blossom has expanded for a time, 
the style with its terminal divisions increases ver}^ considerably 
in length. These latter also separate from each other, and ex- 
pose tlie mature stigmatic surfaces. The style, as it grows, is 
forced by the hood to lengthen in a curve, and thus the stigmas 
are brought into such a position that they block up the entrance 
into the mouth of the corolla, just as the lower anther cells 
block up its throat (fig. 8). A bee v.diich visits the flower in 
this condition cannot possibly enter without rubbing its back 
against the projecting stigmatic surface, so that it cannot fail 
